Meetings of the London-based Player Piano Group as a rule are fairly
jolly and informal affairs, as something of a reaction against the
common format 20 years ago when an owner might spend an hour or more
demonstrating his (it's usually his) instruments before inviting
attenders to try them out.  Nowadays maybe you'll get no more than
an apologetic few minutes of self-introduction -- and with some PPG
notables you don't get a formal opening at all.

This trend is not very instructive for those relatively new to the
field and Denis Hall made PPG history on January 10th by holding a
seminar no less than two and a half hours long, with a snack break in
the middle and a meal at the end.  Title: Duo-Art -- Musical Instrument
or Clever Collectable ?  (Some members had touched Denis off by sug-
gesting that reproducing pianos may play excellent music, but they
don't reproduce.)

Luckily I have his program by me to remind me of the proceedings.
Denis is an expert on early recorded sound and if there is a surviving
disc recording comparable to one of the same piece on roll, he will
have access to it.

To start with, he played a selection of rolls starting with the very
earliest, showing how Aeolian Co slowly got the grasp of putting a
piano performance on paper.  The first series (example: 5501 Love's
Last Word, pl. Arndt) were simply hand-played Metro-Arts of c1912 with
coding added and the artist giving rather more help than was usual
later on.  As these were mostly salon pieces, the treatment wasn't
too critical.

Then (c1914) H Creary Woods, the chief editor, developed the roll-
coding console, a dark and gloomy affair with two large knobs with
pointers on dials and two foot pedals.  Before the performance he
would have invited the artist to mark the score up to indicate mood and
accents.  This was propped up on the console to one side and while the
electrical contacts on the piano made sure that note durations and
pedalling were accurately registered on a master roll being cut in
real time, contacts on the console registered accompaniment and theme
dynamic codes.

What the pedals did is something of a mystery but it is now believed
that one provided additional sustaining to that provided by the artist
and the other commanded the perforator to regard certain notes as
"theme" notes.  After some practice Woods realised that emoting on
the knobs in sympathy with the music didn't produce as good a result
as learning what knob movements produced the same kind of result as the
pianist.  Long hours were spent re-editing and the resulting rolls
(5568 Chopin, Waltz Op 64/2, pl. Ariani) were a distinct improvement.
Demonstrations began to rope in well-known pianists.

About 200 rolls into the list (c1916) (5736 Liszt's Waldesrauschen,
pl.  Donohue) someone blew a fuse at the cost of editing and the rolls
tended to dispense with the many small changes in accompaniment level
which the console produced at the first stage.  The effect is smoother,
but rather bland.

Then (c1919, 6117 Stanford, A Reel, pl. Grainger) Hofmann and Grainger
took an interest in editing their own rolls to exhaustion and a great
deal was learnt about the dynamic shape that coding should take in
certain kinds of piece.  It was found that involving more than one
editor produced better results.  Several stages of prototype roll
survive at the piano archive in Maryland showing that a piece could
undergo up to five ever-finer re-codings.  These rolls ended up (6330
Bach, Adagio, pl. Cortot) being truly convincing representations of
the original performance.

To demonstrate this, Denis then played, on tape, comparative excerpts
from three near-contemporary pairings of disc recording and Duo-Art
roll (played on his own two grands) dating from after 1921.  With some
of these he was morally certain the editor and artist had listened
to the disc when tweaking the roll, but with one (6926 Gabrilowitsch,
Melodie in E minor, pl. composer) the disc was never issued and yet the
performances were still close.

These pairings were, needless to say, highly convincing -- even as
to the tone of the pianos -- and one wished for a Devil's Advocate to
whisper how hard Denis had looked for the good bits.  He closed, after
the break, with an impressive recital of ten rolls, remarking that
where discs exist of the artist playing the piece, these too are close.
He has seen no evidence of discs being recorded at the time of roll
recording.

Luckily I was sitting next to one of the strongest critics of repro-
ducing pianos and so received the alternative view that enables one
to look beyond the immediate impression.  "They're wonderfully clever
machines," he said, "and of course if you're in 1926 it's damned close.
But listen to it carefully.  It doesn't _breathe_ somehow.  On the
roll, the playing sounds stiff and preoccupied.  Considering the
circles they had to run in to do the coding, is it surprising ?"

My view is that you have to listen to the poor rolls first to see
what he's talking about.  Then you can detect the effect, though very
slightly, in the better ones.  What is it ?  I suspect it's tiny shifts
of level in the accompaniment that add to the mood and are too small
for the editor to pick up consciously.  I meant to ask Denis, "what
does study of velocity levels in well-played MIDI performances tell
us about accompaniment coding ?"  but the pizza was delivered and the
seminar ended with a thunder of feet.  (And I'm not sure that Denis is
a great MIDI fanatic anyway.)

This event may have been a bit high-powered for some participants
but for me it was a very rare privilege.  You don't often get two near-
perfect Duo-Art grands, superb hi-fi, a collection of contemporary disc
recordings and an expert like Denis all in one room at once.  What on
earth would the original makers of the Duo-Art have thought ?  I'll
tell you what I think:  "Wow ! Those pianos are _new_!  How do they do
that in 70 years' time with Aeolian dead and gone ?  And that record
player !  Unbelievable."  And, then, after a bit of thought -- "He did
us proud, don't you think ?"

I think he did.

Dan Wilson, London